Jorick van Weelie
Marketing Lead at DataNorth | AI Enthusiast & Tech Storyteller
Articles written by Jorick van Weelie
AI Watermarks: Text, images, and the push for provenance
AI watermarking is now required in the EU. In this article we explain how image and text watermarks work, where each one breaks, and what it means for your content.
Z.ai releases GLM-5.3-Flash
Z.ai released GLM-5.3-Flash on 26 August 2026. The open-weight multimodal model contains 320 billion parameters but activates 18 billion per token. Z.ai positions it as a lower-cost alternative for coding,
Alibaba releases Qwen3.8-Flash-Next
Alibaba released Qwen3.8-Flash-Next on 26 August 2026, an open-weight multimodal model that previews Qwen4 architecture. It activates only 6 billion main-model parameters per token and supports 262,144 tokens natively, with
Google launches Gemini 3.5 Transcribe
Google launched Gemini 3.5 Transcribe on 26 August 2026, a speech-to-text model with a 2.6 percent word error rate on recorded audio and 4.0 percent on live streams. It detects
Skild AI launches S1: a robot model that learns 10-minute
Skild AI announced S1 on 25 August 2026: a robot foundation model learning unseen 10-minute tasks from one human video, at 66% success versus 9%.
IBM releases Granite 4.2
IBM released Granite 4.2 on 25 August 2026: open 3B, 8B and 30B reasoning models under Apache 2.0, with 57.00 on SWE-bench Verified for the 30B.
Alibaba launches Wan3.0 video model
Alibaba released Wan3.0 on 24 August 2026: 30-second 1080p video from text, PDFs and slides, priced at $0.20 per second, plus open SearchQwen models.
DeepSeek Launches V4-Flash-Vision-Exp
DeepSeek released V4-Flash-Vision-Exp on August 21, 2026, an experimental vision model that beats Claude Opus 4.8 on 3 of 11 agent benchmarks.
AI grants and funding in Europe: what your company can
European companies can fund AI through EU grants, free compute and advice, national R&D tax credits and EIB-backed loans. Almost none of it pays for buying AI; it pays for
Liquid AI releases LFM2.5-DSpark draft models
Liquid AI released LFM2.5-DSpark on 20 August 2026: three draft models giving up to 3.18x faster decoding on LFM2.5, with identical output.
xAI’s Grok 4.6 reaches Amazon Bedrock
Grok 4.6 went generally available on Amazon Bedrock on 19 August 2026 at $2 and $6 per million tokens, with a US-only inference profile.
Ornith releases Ornith-1.5
Ornith released Ornith-1.5 on 19 August 2026 in 397B, 35B and 9B sizes under MIT, scoring 86.1 on Terminal-Bench 2.1. Weights on Hugging Face.
The AI assessment: What it delivers, and what it costs
An AI assessment examines where AI can take over work in your organisation and ends in a scored list of opportunities, a shortlist with business cases, and a roadmap. What
Tencent releases UI-Mate-27B
Tencent released UI-Mate-27B on 17 August 2026, an Apache 2.0 open-weight computer use agent scoring 77.0 on OSWorld-Verified, above Kimi-2.6.
Z.ai releases GLM-5.3
GLM-5.3 is Z.ai's post-training only upgrade to GLM-5.2, released 14 August 2026, with a 1,000,000 token context window, 28.3 percent on Terminal-Bench 3.0 and 84.5 percent on CyberGym on Z.ai's
Dots studio releases dots3-note Preview
dots3-note Preview is Xiaohongshu's first public dots3 model, released 14 August 2026 under Apache 2.0, with 280 billion total and 16 billion active parameters, a 512,000 token context window and
Alibaba releases Qwen3.8-27B
Qwen3.8-27B is Alibaba's 27B dense multimodal open weights model, released 14 August 2026 under Apache 2.0, with a 262,144 token context window and 61.7 percent on SWE-bench Pro on Alibaba's
DeepSeek releases V4-Pro-0813 and open sources Harness v0.1
DeepSeek-V4-Pro-0813 is the production release of DeepSeek's flagship model, available since 13 August 2026 with a one million token context window, 87.9 on Terminal-Bench 2.1 on DeepSeek's own harness, alongside
Google releases Gemini 3.7 Flash
Gemini 3.7 Flash is Google's new low cost coding and agent model, released 13 August 2026, with a 1,048,576 token context window, 65.3 percent on DeepSWE v1.1 on Google's own
xAI releases Grok 4.6
Grok 4.6 is xAI's new flagship large language model, released 12 August 2026, scoring 61 on the Artificial Analysis Intelligence Index and priced from 2 dollars per million input tokens
